An automotive parts distribution platform serving importers and dealers across multiple markets needed rapid feature delivery but carried a database design flaw from its early stages. Azati joined as a core engineering partner, resolved the architectural issue through a structured ADR process, and has been continuously extending the platform ever since.
PHP engineers embedded full-time in a multi-vendor delivery team
platform supporting dealers across multiple countries and currencies
external systems integrated: vendor APIs, tracking, PDF generation, and SQS
Automotive parts distribution is deceptively complex. On the surface it looks like an ordering system: a dealer requests a part, an importer sources it, it ships. But underneath sits a web of credit terms, warranty cases, multi-currency pricing, regional tax rules, vendor-specific ordering APIs, and tracking integrations that all have to work together without breaking when a new market is added or a supplier changes their API contract.
The client operates a B2B distribution platform connecting automotive importers and dealers across multiple markets, with dealers in Poland as the primary audience. The platform was already live and growing when Azati joined, but a database design decision made early in the project's life was now blocking an entire product epic. Feature delivery was falling behind. Azati was brought in to stabilize the architecture and keep the platform moving forward.
The original database schema had been built without accounting for one of the platform's core feature sets. By the time the gap surfaced, the system was already in production with real data. A redesign had to be designed, justified to the CTO, and implemented without breaking what was already running:
The platform serves markets with different languages, currencies, tax regimes, and regulatory requirements. Every new feature has to work correctly across all of them, and the business logic for things like warranty handling and credit orders varies by region:
The platform's value to dealers depends on live connectivity to supplier systems. Automated parts ordering, order confirmation, and shipment tracking all depend on external APIs that each have their own contracts, authentication models, and failure modes:
The engineering team is made up of developers from multiple vendors, all working on the same codebase simultaneously. High commit volume, AI-assisted development, and no formal release process creates a coordination challenge that requires discipline from every contributor:
The client needed engineers who could integrate quickly into an active multi-vendor team, navigate a sensitive architectural refactoring without disrupting delivery, and contribute meaningfully to both backend and frontend across a broad feature scope:
When the database design flaw surfaced, Azati's response was a formal written proposal before a single line of schema change. The problem, the options, the rationale, documented, reviewed by the CTO, approved, then implemented. That sequence matters in a multi-vendor environment where undocumented decisions create compounding risk.
The project requires fluency across Laravel backend development, Angular and TypeScript frontend work, PostgreSQL schema design, AWS infrastructure, and external API integration. Azati's engineers contribute across all of these without needing separate specialists for each layer, which matters in a team where coordination overhead is already high.
The team runs AI tooling as a standard part of daily engineering, not as an experiment. Azati's engineers operate the same way: AI-assisted code generation and analysis, Playwright MCP for browser automation in testing, structured prompting against documented APIs. Joining a team already working this way means no friction, no ramp-up on tooling culture.
Two engineers, full-time, same codebase, same responsibilities as any other developer on the project. Not a separate workstream that delivers on a schedule, a genuine part of the team's daily capacity.
If your automotive platform is falling behind on features, carrying an architectural debt, or needs engineers who can ship across the full stack without slowing the team down, let's talk.
Fix my platform architectureThe foundational work of the engagement was resolving a database design issue that was blocking an entire product epic. Azati produced a formal Architecture Decision Record laying out the problem, evaluated options, and recommended approach. After CTO approval, the schema was redesigned and the migration implemented without interrupting the ongoing delivery cadence of the wider team.
The core of the platform is the order lifecycle between importers and dealers. Azati contributed to building and extending order management features including visual order status tracking, parts catalog management, price request flows, and dealer credit order handling. The system handles the full importer-to-dealer supply chain with the regional and currency variations each market requires.
The platform's operational value depends on live connectivity to external systems. Azati built and maintained integrations with vendor ordering APIs for automated parts procurement, order confirmation flows, and shipment tracking systems. Amazon SQS handles async messaging between the platform and external services, ensuring reliability under variable load.
The platform generates invoices, shipping documents, and other structured outputs using a third-party PDF generation API with template-based rendering. This allows the business to produce consistent, branded documents across different order types and markets without maintaining a custom PDF rendering stack.
The Angular and TypeScript frontend is built against a well-structured API layer documented via OpenAPI and Swagger. Azati contributed to frontend feature development while helping maintain the API contract discipline that makes the codebase tractable for a large multi-vendor team, where consistent interfaces reduce the coordination overhead of many engineers working on the same system simultaneously.
| Metric / area | Before Azati's involvement | After Azati's involvement |
|---|---|---|
| Database architecture | Schema blocking core product epic | ADR approved by CTO, schema redesigned and migrated |
| Epic delivery | Blocked by architectural constraint | Epic unblocked and delivered |
| Order management | Basic order flows | Extended with status tracking, price requests, credit handling |
| Vendor connectivity | Manual or limited integration | Automated ordering and confirmation via vendor APIs |
| Document generation | No structured output pipeline | PDF invoices and documents generated via API |
| API structure | Inconsistent contracts | OpenAPI/Swagger documented, generator-ready for frontend |
All platform traffic is secured with TLS 1.3 encryption across client-server and service-to-service communication. File storage uses Amazon S3 with access policies appropriate to document sensitivity. Async messaging between the platform and external systems runs through Amazon SQS with standard AWS access control. The platform handles multi-region financial data including credit orders and invoices, and operates under the data handling requirements applicable to B2B commerce platforms serving EU markets.
Two Azati engineers work on a Time and Material basis, giving the client flexible capacity that scales with the backlog rather than a fixed project scope. Hours are tracked transparently and billed against actual delivery.
The team operates on a Kanban basis with a continuously flowing backlog rather than fixed sprints. Features are developed, tested, and shipped in a continuous cycle:
The schema issue that was holding back an entire product epic is gone. The redesigned database now supports the full planned feature set, and the product roadmap is no longer constrained by an early architectural mistake.
Azati engineers have contributed continuously to platform feature development across order management, vendor integrations, document generation, and frontend work, keeping pace with a fast-moving product backlog.
Every feature delivered works across the existing language, currency, and market configuration. Expanding to a new EU market doesn't mean starting over it means turning on what's already built.
The investment in OpenAPI documentation, Swagger-generated frontend service clients, and structured data contracts means the team's AI-assisted development workflows operate at high efficiency. The codebase is legible to AI tooling in a way that most codebases are not.
The engagement has built genuine expertise inside Azati in automotive B2B distribution mechanics: importer and dealer roles, credit ordering, warranty handling, cross-region tax and currency logic, and the vendor API landscape for parts procurement.
Two Azati engineers work full-time as embedded members of the client's multi-vendor development team, contributing across the full stack from database architecture to frontend feature delivery.
Adobe AEM support and full DAM migration for a global beverage enterprise: 10,000+ assets transferred with 100% metadata integrity in under 2 months.
Offline computer vision system for automated pallet counting using edge AI.
Modernized retail corporate system with optimized warehouse processes and enhanced efficiency.
Modernized warehouse management platform rebuilt on modern tech stack for scalability.
Modernized Shopify app with enhanced stability, performance, and engagement.
Web platform for drone ordering, rental, and repair services.
Last updated